A network system is defined as a group of nodes which are interconnected in a pre-determined fashion. The nodes may work in a synchronous manner, in an asynchronous manner, in a collaborative manner, in a distinctive manner, in a sequential manner, in a non-sequential manner, in a discrete manner, in a pattern, out of a pattern, or in any type of manner predetermined by the intended use of the network.
In networking, load refers to the amount of data (traffic) being carried by a network. One of the parameters of an efficient system, especially, of a network system is the amount of load that it handles.
For many loads; whether physical or virtual, load distribution is an important function of network attributes. Load changes may be in the form of load addition, load deletion, or mere load change. This comprises resource additional, resource deletion, resource changes, node addition, node deletion, node changes, data addition, data deletion, data changes, and/or the like.
Network Load Balancing (NLB) is the management of traffic across a network.
In at least a first embodiment, Network Load Balancing distributes workload across multiple CPUs, disk drives and other resources in an effort to use network resources more efficiently and avoid network overload. Load balancing may be accomplished through software or hardware.
Network load balancing is an efficient and cost-effective solution designed to enhance availability and scalability of Internet applications by allowing system administrators to build clusters, which are load balanced with incoming client requests.
In at least a second embodiment, a load balancer is a device that acts as a reverse proxy and distributes network or application traffic across a number of servers. Load balancers are used to increase capacity (concurrent users) and reliability of applications. They improve the overall performance of applications by decreasing the burden on servers associated with managing and maintaining application and network sessions, as well as by performing application-specific tasks.
In at least a third embodiment, Network Load Balancing distributes IP traffic across multiple cluster hosts. It also ensures high availability by detecting host failures and automatically redistributing traffic to the surviving hosts. Network Load Balancing provides remote controllability and supports rolling upgrades.
Typically, Network Load Balancing enables a network to deliver high performance and failover protection
Network Load Balancing servers (also called hosts) in a cluster communicate among themselves to provide key benefits, including scalability and high availability (of nodes).
In at least a fourth embodiment, Network Load Balancing uses port rules to customize load balancing. Port rules can select either multiple-host or single-host load-balancing policies. Port rules also comprise handling priority.
According to the prior art, one of the techniques for load distribution is a DHT (Distributed Hash Table) technique. This technique considers logical distance between the nodes. This technique makes a choice of destination node to transfer the extra load on the basis of logical distance analysis, and not on physical distance measure.
According to the prior art, one of the techniques for load distribution is a Static Load Balancing technique. This technique is not suitable for distributed applications.
According to the prior art, one of the techniques for load distribution Dynamic Load Balancing technique. This technique is not suitable when the system in stable enough.
Furthermore, according to the prior art, all prior art technologies are not scalable in a co-ordinated manner.
FIG. 1 illustrates one of the existing models of grid layout. Below is an associated Table 1.
TABLE 1Number ofNo. ofTimeresourcesresourcesInstanceavailableremovedGrid-1Grid-2Grid-3Grid-4T132 * 4 = 128032323232
FIG. 2 illustrates one more of the existing models of grid layout, where resources and nodes have been removed. Below is an associated Table 2.
TABLE 2Number ofNo. ofTimeresourcesresourcesInstanceavailableremovedGrid-1Grid-2Grid-3Grid-4T232 * 4 = 1283032 − 5 = 2832 − 7 = 2832 − 10 = 2232 − 8 = 24